The Designer's Guide Community Forum

The Designer's Guide Community Forum
https://designers-guide.org/forum/YaBB.pl
Simulators >> RF Simulators >> noise simulation using port as input (cadence )
https://designers-guide.org/forum/YaBB.pl?num=1194006190

Message started by Julian18 on Nov 2^nd, 2007, 5:23am

Title: noise simulation using port as input (cadence )
Post by Julian18 on Nov 2^nd, 2007, 5:23am

Hi, there
I am simulating the NF of a differential LNA (tool: cadence, spectre), and i found the setup confused me much. I use port as my input source,connect in parallel a resistor with this port and assign to this resistor the same resistance value as that of the port resistance, under this set up I can get a NF curve. Of course if I drop the resistor in parallel with the port, I got a different curve. surely there is only one true curve. so which setup should I take to get the right value?

B.R.

Julian

Title: Re: noise simulation using port as input (cadence
Post by pancho_hideboo on Nov 3^rd, 2007, 4:47am

Quote:

I use port as my input source,connect in parallel a resistor with this port
and assign to this resistor the same resistance value as that of the port resistance,
under this set up I can get a NF curve.

Why do you connect resistance in parallel with port ?

Quote:

Of course if I drop the resistor in parallel with the port, I got a different curve.
surely there is only one true curve. so which setup should I take to get the right value?

If parallel resistance is one of part of DUT, for example, you are aiming broadband resistive matching,
former is correct.
If not, latter is correct.

Title: Re: noise simulation using port as input (cadence
Post by Julian18 on Nov 3^rd, 2007, 7:45am

pancho_hideboo wrote on Nov 3^rd, 2007, 4:47am:

Quote:

Why do you connect resistance in parallel with port ?

since in cadence if you specify ac=1 the actual ac value is 2, so i guess that in doing so I can also get the right value in noise simulation, but I am not sure.

B.R.

Title: Re: noise simulation using port as input (cadence
Post by pancho_hideboo on Nov 3^rd, 2007, 11:06pm

Quote:

since in cadence if you specify ac=1 the actual ac value is 2,

This is no problem as far as you calculate appropriate gain by your own definition and method.
This mean that you must not use direct plot in ADE without understanding gain definitions of direct plot.
Even if you connect same value resistor with port, voltage at DUT input is not 1volt, since DUT input impeadnce is not infinite.

Quote:

so i guess that in doing so I can also get the right value in noise simulation, but I am not sure.

What do you mean by "the right value in noise simulation" ?
What is your judgement, regarding right or not ?

Assume simulation of LNA.
If you connect matched resistor in parallel with port, you can never get NF lesser than 3dB